Recently, I've noted that many people in Israel, thinks that if an appliance causes the residual-current device (RCD) to trip, this means that the appliance is short circuiting, and inviting a technician or an electrician to search which part of the appliance is shorting.
This, despite RCD isn't a circuit breaker. RCD isn't intended to protect the building from a short circuit or an overloading.
Most israeli people also recognises the RCD as the circuit breaker and confuses with them.

Well, the part about an RCD tripping being a result of a short circuit in apliances like fridge or heaters I would see as a truth: Most likely the short circuit happen between some live wiring and the grounded schassis.
The big contributorto that is the fact, than the schassis have no insulation (beside the lacquer), so when some wire rubs over some edge, it most likely end up as a short between this wire and the metal edge.

And in heaters the short circuit most frequently happen between the heating wire and the tube, when the tube crack and a moisture seeps in.

And the fact people do mix RCD's vs breakers is of no mystery (and I would even say no real problem): When the applince is in good shape, it does not trigger any of them. If some of the devices trip, there is a fault inside the appliance, so the appliance need a qualified repair and/or replacement, period (for non-electrician people).

I even think majority of installations would desire the usage of only RCD-CB combo's on each branch, as the ground-short faults are even more common than line-neutral faults, what with single, common RCD mean your house remain without any light... With the combo, only the affected branch would be shut down and for the user it does not matter why exactly it shut down, the important thing is, there is something unsafe, period. To resolve what is then the job of the qualified technician.

Medved put it up very well

Where this is a problem, is this :

1. Users see the overcurrent breakers and assume that they have intact RCD protection, while actually there is none

2. This makes for the electrician for harder time to troubleshoot the cause of the problem



Medved, i opened a couple of those "combined breakers" and really ? im by far not impressed

The stand alone RCDs are entirely magnetic inside - a current transformer with secondary, and its secondary is powering a solenoid to trip it. No single "electronic" component in there (as long as you dont count the resistor for the test button)

The combined ones have some electronics in them, which are IMHO a whole new world for unreliability - All the known problems with deriving power for such circuits from the mains (all the best with all its problems - capacitive droper vs zener). bad clearance between components, and if this is not enough, the circuit is placed aftre the RCD (to disconnect itself when it trips), i can imagine what will remain of it if it trips while there is some moderate inductive load connected on the branch.....

The standard magnetic RCD for me every day

The electronic is required for all RCD's with a sensitivity of 30mA and below, if you don't want false triggering by capacitive currents.
And the 30mA is, according to the recent code,  the maximum fault current for the RDC to trip. The reason is for the deviceto trip even with a resistive leakage, e.g. when somebody staying on a grass touches the line potential. The limit is set for the current level, when there become a risk of cramp preventing you from releasing the electrified object and at the same time not to cause false triggering.

The basic concept of an electromagnet and current transformer respond to current of any phase, so when the sensitivity is set at the 30mA.
But the EMC filters in many appliances have capacitors between working wires and the PE, what add up to the leakage current. When multiple of such appliances are connected, or the wiring is long (e.g. garden wiring), the normal, capacitive current already approaches the 30mA threshold, so prebias the triger magnet. The consequence would be too frequent false triggers, yielding "experts" to bypass the RCD...

As the fault current is resistive, what matter is only the resistive part of the current. So if the device could be made sensitive to only resistive component (= in phase of the voltage) of the fault currentand insensitive to the reactive components (thse caused by the capacitors and long wiring), the threshold of 30mA and even below could be fulfilled and at the same time the false alarms avoided.

And it is only the electronic, what make this filtering feasible (both technically and economically). And as the integration of the RCD with a circuit braker is only a recent "invention" (it required to solve the problem of high current breaking capability contacts in a small space), these newer devices already came with the phase filter.

The standards here require 30mA RCD for all residential and office installations. So far i only seen them trip when loaded up with a clasroom of computers (eg at school) and only when flipping on the main (so all inrush of all the Y caps at once), Never seen "average" amounts of IT equipment cause nuisance tripping even when it is in the moment of connecting all at obnce. So for "general purpose" systems (where the combined breaker is supposed to go) look like the problem of nuisance tripping does not exist

Same for outdoor systems, i seenthem trip 30mA RCD but only in times whern there is real fault, i used several times on the order of 150 meters of extension cables (plugged into each other) with no effect on a 30mA RCD, and this RCD works

The 30mA is just on the border: Small installation do suffice without the filter, but the setup is more sensitive (main trouble makers are the washers and/or fridges, the computers have generally way lower value "Y" capacitors)
The standard for 30mA sensitivity is not here for so long (measured in lifetime of the installation), are you sure the non-electronic RCD was the 30mA and the one not triggering with larger capacitive "load"?
Because since the introduction, the RCD's were sold in two categories: One "standard" (pure electromagnetic) and one "more immune"/"for large installation" (electronic).
The experience tells, than the first ones tend to trip even on some mains disturbance and their sensitivity is really close to the 30mA limit.
The "immune" group really need resistive current to trip and for that their sensitivity (for the resistive component) is more around 10mA, so provide quite better degree of protection against real faults (although according to the standard they still belong to the 30mA category).

And to avoid misunderstanding: All this apply only for single phase RCD's.
Three phase units can not filter the phase, as the real fault should be expected in any of them. But the 3-phase system provide quite balanced capacitive load, so the capacitive current responsible for false triggering is balanced out, so the filtering is in fact not needed.

And using a 3-phase RCD to protect single phase branches is asking for huge troubles: The RCD's opens the Neutral, so it may happen, than the Neutral contact fail and does not close. With single phase RCD there is no problem, only the load is on line potential. But when the Neutral does not close on the 3-phase RCD, some phase branches may get severe overvoltage.

When there are only 3-phase appliances on the 3-phase RCD, the 3-phase appliances usually survive that, as they either do not use the Neutral at all (most saws, pumps,...) or to the Neutral is connected only some control part, so rather low power.

And here again the largest troublemakers are the stove/ovens. But these have only fixed connection (no sockets,...), so do not require the RCD at all (RCD's are required only on sockets, not on fixed installations).

The code here recommend RCDs since the end of 70s and require it since early 90s. The RCD specified was allways the same - 30mA. Except for big groups of IT equipment (and the ones at school were on 3 phase btw) i never seen any false tripping with a plainmagnetic RCD

The 3 phase RCDs heve one contact intended as neutral - it is opening after others and closing before others. As long as it is installed correctly (with the neutral going through this contact and not some other) it won't have any problems of the open neutral type

Stoves need Y capacitors ? Since when ? The stove is just a resistor.....

Stoves need Y capacitors ? Since when ? The stove is just a resistor.....

Capacitors do not need stoves (except induction ones), I mean appliances as washers and so on (with brush motors or with speed control and/or frequent automatic switching)

The stoves I mentionned as troublesome as they are three phase appliances, but load individual phases unevenly, so the open neutral could cause them to become damaged even when alone on that branch.

And for the Neutral closingfirst/opening last: This is indeed their design, but only of the mechanical part, there is no electrical interlock. So when some dirt enter the contact area, it could prevent the ontact from closing, even when the contacts are already pushed together.
And this is quite frequent problem with RCD's (as they get regularly open by the monthly tests;the CB's don't suffer from this, as they normally remain closed all the time).
For the 3-phase RCD's it would be beneficial to have an open neutral trigger (so it would trip and open, when the Neutral does not really electrically close), but I do not have any idea, how to implement that in a simple, but still reliable way...

Double contact for the neutral. The aux one is small (meant to carry small current), connected through a resistor (to prevent some of the normal current going through it), and bypass the current transformer. Normally the current in the neutral will avoid the resistor so go as usual. If there is bad contact some of the neutral current will go the alternative way through the resistor straight to the primary neutral, bypassing the current transformer and trip the RCD. Resistor is calculated according to max allowed dV between primary and secondary neutrals / 30mA. Power that the resistor must withstand continuously also derived from there

Maybe not a resistor, but simply a separate voltage winding on the trigger electromagnet. In that way the residual current could be still kept safe, when the aux contact does not open and there would be some leakage through the fault in the installation.

But I have not heard about any commercial RCD with such feature. At least for the stoves it would be quite a help. Thereare faults yielding quite severe fire risks, when the stove is not on a RCD (short circuit of the heater to the outer tube in the middle of the heater, when only one side is switched - I've seen few such ovens to really melt the heater before the current rise to trip the CB...)